Tuesday, November 21, 2023

UN 165: Bus upgrades for a broke government


I'm getting the same message from everyone in government circles: "we're broke". 

But that shouldn't mean they should give up on improving bus services. 

Indeed buses are the best option if you want to improve transport for a lot of people in a lot of suburbs. Especially with the opportunities open due to Melbourne's historic slowness with bus reform, with many decades-old inefficiencies and complexities remaining to be fixed.    

Back in March I suggested ten super cheap bus boosts for 2023. The good news is that one of those ten improvements got done with the 271 gaining Sunday service last month

That wasn't so hard was it?

The need to identify cost-effective bus improvements has got even more pressing in the last year with rising concerns over cost of living pressures and housing affordability. And it's not just individuals; governments that borrowed heavily are also feeling the pinch with interest rates soaring.



Although there's still cranes in the sky, it feels like we're in a 'major project autumn' with few if any new commitments being made and some older ones never actioned (Rowville tram anyone?). However we're still growing, so the case for continued transport improvements, especially the type that is either cheap or works existing assets harder, is overwhelming. That basically means a mix of active transport links, off-peak rail frequency upgrades and big dollops of bus reform. 

Unfortunately Victoria's Bus Plan, intended to tackle the latter, is the stunted child in the government's transport agenda. Buses never got the largesse that road and rail infrastructure did in the boom years. There wasn't much new in the 2023 state budget either.

While Infrastructure Victoria publicly promotes bus reform, DTP's Paul Younis & co are strangely quiet despite a plan existing. For example they chose not to give Victoria's Bus Plan a single mention in DTP's recent annual report. Bus Plan has also apparently vanished from the DTP website with nothing searchable at the time of writing. Then minister Ben Carroll vigorously promoted the benefits of improved bus services but presumably too few others in government agreed to make it a serious priority (especially if they thought transport was already generously funded, albeit for infrastructure, not services). When you add these points it's hard to escape concluding that the Bus Plan currently has only a low status within DTP and government. One hopes new minister Gabrielle Williams can revive departmental interest and win support within government for bus reform and funding. 

Victoria's Bus Plan was essentially a 'plan for a plan'. Many details that one might have expected in it got held over for the Bus Reform Implementation Plan. That's not out yet. It may still meet its late 2023 deadline. However almost 30 months to wait is way too long. We've removed level crossings in less time than that. One can't help thinking whether lead times would be shorter if DTP's top executives included more Ken Mathers-like figures who get stuff done. 

The late 2022 announcement to review bus networks in Melbourne's north and north-east was welcome. However the review's size involving over 100 often interdependent bus routes likely extended time-lines, especially without certainty of funding. While well-intended this approach may not suit current budget circumstances nor the need to build delivery capability and momentum with early and closely spaced 'quick wins'.

I'd have prioritised the latter, with a larger number of simpler initiatives done quickly and concurrently. Instead of starting at the top, DTP could have started at the bottom of the 'bus reform pyramid' below. And high  patronage potential routes in areas like Dandenong should have been improved first to maximise early benefits. 



As DTP reform and delivery capability grows work could move up the pyramid, implementing more advanced reforms. Work in several areas should proceed in parallel to spread benefits as widely as possible.

I say this because of the experience about 10 years ago under PTV. In 2014 it implemented significant bus network reform in Brimbank under a minister receptive to bus reform. Unfortunately the opportunity to do likewise in the east, eg around demographically similar Dandenong, was not taken despite similar social needs and network issues. Later the politics changed and the door to bus reform closed. Hence routes and timetables in areas like Noble Park North and Dandenong North today remain inconvenient and unreformed now because PTV was not agile enough to act when the opportunity existed.     

What are some specific 'base of pyramid' reforms we could be thinking about today? Here's such a list, refined from the March item with more detail added and offsetting savings identified: 

Timetable changes only

These redistribute bus service kilometres from low to higher patronage potential routes, benefiting more people. They are the simplest to do as they are within the same bus operator group and do not require new bus purchases.  


1. Route 800 longer operating hours, higher Saturday frequency and new Sunday service. 
Funded from reduced service on much quieter Route 704 and deletion of the largely duplicative Route 698. More here.
Benefits: Would provide 7 day service to Chadstone Shopping Centre along Princes Hwy on Melbourne's most productive bus route without it. 


2. Route 420 Sunday - Thursday evening operating hours extended to midnight approx.
Funded by reducing or eliminating weekend service on Route 422 (with limited unique coverage). 
Benefits: Would extend after 9pm service to a large low-income/high patronage area currently without it. Route already operates 24 hours on weekends as a part of Night Network but finishes approx 9pm other nights. 


3. Route 220 earlier Sunday start. 
Funded by reducing early Saturday frequency on Route 223 from every 15 to every 20 min. 
Benefits: Would increase Sunday span on popular routes with earlier CBD arrivals. Even just adding one trip for a 30 min earlier start would be very worthwhile given the area's demographic catchment.   


4. Route 431 operating hours extended to 9pm and 7 day service. 

Funded by reducing Route 432 from an uneven 20-30 min to an even 30 minutes off-peak weekdays. More here.  
Benefits: Adds evening, 7 day and public holiday service in Kingsville area without it. More reliable off-peak connections with trains and a clockface timetable on the 432. Route 432 currently has very low patronage productivity so some transfer of resources from it is justified and would likely go unnoticed. 


5. Route 237 weekend service added. 

Funded by reducing weeknight frequency and transferring service hours to weekends. Current timetable is weekdays only despite serving apartments on Lorimer St. 
Benefits: Would provide a basic 7 day service to apartments on Lorimer St. 


6. Extended hours on Routes 580 and 582. 

Funded by reducing frequency on long and poorly used routes 578 and 579 and transferring service hours to Route 582 (which currently finishes early) and Route 580 (which has a late Sunday am start).  
Benefits: Would upgrade both routes 580 and 582 to minimum service standards (ie 9pm finish) 7 days per week. Further benefits may be possible if route reforms are brought into scope, eg making 582 bidirectional and extending it to Greensborough. 



Splitting complex routes

Splitting can simplify long and indirect routes into two straighter routes with a new route number introduced for one half. No stops are missed and there need be no time or timetable changes (although these might be desirable).  Desirable splits could include the following:

1. 380 at Ringwood and Croydon. A complex circular route that could be simplified to two Ringwood - Croydon bidirectional routes, one north and the other south. Weekend operating hours extensions are desirable but the split can be done without them. This route serves Maroondah Hospital. 

2. 469 at Airport West. This split at a major shopping centre would simplify a very complex and circuitous route. Route number 466 is available for one of the sections.  

3. 517 at Greensborough. This Northland - St Helena route has a busier western portion and a quieter eastern portion. The entire route operates every 24 minutes on weekdays, not meeting trains every 20 minutes. Splitting the route at Greensborough with the western portion every 20 minutes and the eastern portion every 40 minutes would better match usage with service provision. The eastern portion (Route number 519 suggested) could have its timing optimised to meet trains at Greensborough with departures evenly staggered with the existing 518 (also every 40 min). 

4. 566 at Greensborough. A complex route that backtracks via Greensborough with some stops served by buses in both directions. No one would ride it end to end. Hence it is a good candidate for a split at Greensborough, adding amenity to the rebuilt station and bus interchange. Other potential improvements include a western extension to Epping Plaza and improving its frequency from the current ~23 to 20 minutes to harmonise evenly with trains. 565 is a spare route number suggested for the northern portion.

5. 736 at Glen Waverley. Again no one would ride this end to end as walking would be quicker and there is a train. Glen Waverley is a major centre, interchange and future SRL station. The service would become easier to use if it was split into two routes. Through running and timetables could even remain the same with a route number changing at Glen Waverley. Spare number 739 could be used for the eastern half. 


Removing kinks and deviations

Some bus routes have kinks or deviations that slow through passengers but do little to improve network coverage. Some kinks may add confusion or leave gaps of up to 80 minutes in the timetable if they only operate on some trips. 

Examples of routes with kinks or backtracking that could be removed include 273 (Blackburn North), 279 (Blackburn Station), 504 (indirectness in Fitzroy North), 503 & 510 (near Essendon), 506 (Smith St), 536 (alternating paths) and 833 (Frankston-Dandenong Rd). Others like 555, 556558, 624, 742 and 895 (to name a few) are also complicated. 

Cutting poorly used kinks may free up service kilometres that could fund improvements, even if it's just adding one or two earlier or later trips on a popular route run by the same operator.  

Economical network reform within the one operator group

While network planning is best by region rather than by bus operator, there are still cases where small cost-effective improvements can be made by redistributing service kilometres within a bus operator's network from quieter to busier routes. 

This is least likely where a bus operator has a. only a few routes, b. relatively low service levels, c. few quiet routes and/or d. an existing or recently reformed efficient network (so there are few further efficiencies). 

Conversely prospects for cheap reform are highest for operators that have a. many routes, b. relatively high service levels, many quiet routes, and/or d. an inefficient network with indirect and overlapping routes that hasn't been reformed for years. 

Below is my first cut at graphing this for various Melbourne bus operators. Approximate size is left to right while cost-effective network reform potential is bottom to top. 


At bottom left is Martyrs. Their only regular route is the 683. It's direct, efficient and well used. So you'd leave it as is. To the right of them is McKenzies. They run a few more routes. However their scope for reform is very low since I've factored in the reformed Yarra Valley network starting in a few days. But if I was comparing the current network then McKenzies would be higher up, somewhere near Panorama

Speaking of which, Panorama, in the top left, is the only smallish operator who I've rated as having high reform prospects. Why? Although they run only a few routes, I regard them as the keystone to bus reform in the Eltham / Diamond Valley area, which has a large number of low productivity bus routes. Reform to these could unlock wider benefits that ripple all across north-east Melbourne.

This is because Panorama run the very lowly used but quite highly served (for a semi-rural area) 578 and 579. If you are able to redistribute bus hours resources from these to more densely populated areas like Eltham, Greensborough and Templestowe then there may be an overall patronage gain. One option could involve changing the circular unidirectional Route 582 at Eltham to a bidirectional Eltham - Greensborough route, passing near Montmorency via the 293 alignment. In conjunction with a kilometres neutral swapping the 901 and 902 alignment in the Greensborough area (providing a direct Greensborough - Doncaster SmartBus connection) the 582 extension makes the 293 (run by Kinetic) redundant. All that extra kilometres could be put on the popular 281 to increase its frequency to 15 minutes weekdays and likely 30 minutes weekends with new Sunday service and longer operating hours. This concept would need to be weighed against alternatives (as better options may be available) but illustrates that even a relatively small change can have a benefit across a wide area. 

What about other corners? CDC is alone in the bottom right. This is because it is the rare combination of being a large operator that has had most of its routes redone during the big Williams Landing, Brimbank and Wyndham reforms of 2013 - 2015. Most of its routes are more productive than the Melbourne average yet service levels are often quite low, with 40 minute off-peak headways common. Boosting frequencies of these would be an excellent idea but new funding is required as there are few if any economies to be found in a generally direct and efficient network. The main exceptions are their operations in the east where there are potential low (but likely not zero) cost reform opportunities involving routes like 606, 623 and 624 remaining. 

The top right is the most interesting. Here you've got three big operators with either high service levels, overlapping routes or unreformed networks. Each has slightly different network issues but economies of scale should make aspects fixable. 

Of these Dysons has the least reformed bus network in Melbourne, especially after taking over the Reservoir Bus operations. Everything that's wrong with Melbourne buses can be found on a Dysons service, whether it be the sparse service of the 609, the backtracking of the 556 or 566, the midday reversal of the 558, the weak terminus of the 552, the midday Saturday finish of the 559 or the prevalence of 22 - 25 minute headways that miss trains every 20 minutes. It is possibly for these reasons that the government chose Melbourne's north for its first bus reviews in September 2022 notwithstanding the complexity involved.

Dysons run no SmartBuses so their service levels are not particularly high. However there are portions of their routes with relatively poor catchments and patronage productivity, for example the Diamond Creek end of the 381 or the 517 north of Greensborough. Rationalising these could free up a few service kilometres for routes that need it more. Generally though cost-effective timetable reforms likely also require network reforms making improvements here harder than for other operators' routes. 

I've rated Kinetic's scope for reform as being slightly lower than Dysons or Ventura thanks to significant network reforms in 2014, 2021 and even a little in 2023. However Kinetic runs most of Melbourne's high service SmartBus routes. Sections of these overlap other routes or serve poor catchments. Notable examples include 901 overlaps with 280/282, 309 and 902 and 903 overlaps with 232, 411, 465 and 527. Such overlaps make the orbitals less consistently productive than they should be and impose opportunity costs, particularly with regard to their currently poor weekend frequencies.

Semi-rural areas like Yarrambat get an orbital SmartBus that is arguably overservicing. And while the Greensborough area gets an (arguably generous) two SmartBus orbitals (901 and 902), none provide a frequent connection between the two big centres of Greensborough and Doncaster, leaving the job to the duplicative and infrequent 293 (that nevertheless has good productivity performance). Swapping the 901 and 902 in both the Greensborough and Broadmeadows areas could provide this connection as well as improve directness to Melbourne Airport.

There is a lot of buses and service kilometres tied up in the SmartBus orbital routes. The government has shown it can reform orbitals by recently rerouting the 903 via Deakin University. If it wants a more efficient bus network for the least money it will have to revive discussion about splitting the SmartBus orbitals despite splitting being one of the controversial parts of the aborted 2015 Transdev network

More than any other Melbourne bus operator, Kinetic has a reserve of bus service kilometre resources that could be used to cost-effectively optimise the network. As well as the orbitals mentioned previously these include expensive to run but poorly used and/or duplicative routes like 232 and the 280/282 Manningham Mover. Potential benefits could include new SmartBus routes, much needed weekend service upgrades on the busiest parts of some orbitals, extra coverage in areas without it like around Ringwood East and simplifying the very complex 600/922/923 corridor between St Kilda, Sandringham and Southland.

Ventura is like Dysons in that it doesn't have a lot of really high service routes and most of its territory hasn't had a lot of bus network reform. Nevertheless some of its routes, like 693 and 742 on Ferntree Gully Rd, have significant overlaps with scope for consolidation. The same can be said for growth areas between Berwick and Cranbourne plus simplification around Pakenham. And there are established areas like Bentleigh East, Mulgrave, Keysborough and Dandenong North where it might be possible to simplify networks and boost service frequencies. Weekend services better than the typical 60 minutes would be very desirable but I'm not sure there's a lot of 'fat' to improve more than a few routes within existing resources.    

Area based network reforms across operators

These are the most challenging for the DTP to arrange but are necessary to maximise 'bang for buck' by removing inefficiencies to deliver the simple frequent and direct services that the Bus Plan (correctly) says we want. The tougher the budget the more important such planning creativity becomes in improving bus services. And there have been local examples of multi-operator sharing for an overall network benefit, such as with the successful Route 900 SmartBus between Caulfield and Rowville. 

A few opportunities for network reform involving two or more operators might include: 

* Ballarat Rd frequent bus. Network reform based on reforming 220 and 410 between Footscray and Sunshine, with 220 operating along more of Ballarat Rd to provide a simpler service linking both VU campuses. Service would be tram-like, ideally every 10 min or better all week.   


* Footscray - VU - Highpoint.
Based on merging 223 and 406 to provide a simple frequent 7 day bus between major destinations. Current services are less frequent than they should be and, in the case of 406, indirect as well. Route 409 may also need reform in the area to retain coverage. Again you'd be wanting a 10 min 7 day service with long operating hours. 


* SmartBus routed via Highpoint.
Based on rerouting the 903 between Essendon and Sunshine via Highpoint, replacing 468 and 408 in area. Would likely require other compensatory reforms, eg longer hours and more frequent Sunday service for the 465 and an extension of the 406 to Sunshine to replace the 903. This mini-review would remove the 903/465 duplication and bring orbital SmartBus services to Highpoint (as intended with the aborted Blue Orbital). 


* Millers Rd Altona North.
Corridor currently has multiple overlapping routes (232, 411, 903) yet still has low frequencies, especially on weekends. Potentially ripe for consolidating with 411 becoming the main frequent SmartBus type route in the area and the 903 finishing at Sunshine. Best done in conjunction with an Altona North network review that features improved and more direct connectivity to Newport Station and potentially Fishermans Bend. 


* Coburg - Heidelberg via Northland.
 Based on merging portions of Route 527 with 903 to provide a 7 day orbital service running every 10-15 minutes between major destinations, trams and trains. More.


* Berwick - Cranbourne.
Based on simplifying and joining local routes to provide a one-seat ride between these key destinations, preferably every 20 minutes or better. Ventura and Cranbourne Transit currently operate in the area with some routes often having large overlaps, low frequencies and weak termini. As Clyde Rd is a growth area corridor an examination of whether current GAIC funding arrangements are flexible enough to permit the most efficient and economical bus network planning would be essential.   

 

Conclusion

A broke government can still make bus reforms. Indeed wise spending requires it. Several ideas to maximise the benefit from our bus assets and funded service kilometres have been presented. While an effort has been made to identify offsetting costs for operations, there will still be some setup costs. However these are worthwhile to kick-start the benefits listed and are small relative to what is spent on major projects or even DTP executive salaries. 

PS: An index to other useful networks is here.

2 comments:

Anonymous said...

One very simple, low cost improvement to bus services is to think about stop access. I was on a major stop on blackburn road today, served by three lines, yet to reach the opposite side meant crossing 4 lanes of traffic with no assistance.

To reach a crossing meant a 1km+ round trip, and of course waiting for the lights to grant you a timeslot. Once lights are considered, it is adding extra 15-20 minutes to people's commute, and more if you are infirm or travelling with children.

In parts of melbourne, this has become much worse, with stops moved from local roads, with some chance to cross, to high speed, multi-lane roads that could for all intents and purposes be considered a brick wall between the two stops.

Anonymous said...

One easy win for bus services would be to prohibit parking in bus lanes, another would be to stop cars using the bus lane to queue for left turns.

Another problem with busses, when they are forced to enter shopping centres, they have no priority over cars and get stuck in traffic. This needlessly slows down bus timetables, most of all at peak times.